Data sharing system and data sharing method therefor

A method for machine creation, comprising

• inputting a random noise and reference data into a memory;
• transmitting, by a computation device, an input data of the random noise into a generator neural network to perform operation to obtain a noise generation result;
• inputting, by the computation device, the noise generation result and the reference data into a discriminator neural network to obtain a discrimination result;
• updating, by the computation device, discriminator neural network parameters and generator neural network parameters according to the discrimination result,
• wherein the updating, by the computation device, the discriminator neural network parameters and the generator neural network parameters according to the discrimination result includes:
• computing loss values of a generator neural network and a discriminator neural network respectively according to the discrimination result;
• adaptively updating a parameter in the discriminator neural network according to maximum gradient directions of reduction of the loss values to further improve discrimination accuracy of a discriminator;
• adaptively updating a parameter in the generator neural network according to maximum gradient directions, discriminated by the discriminator, of increase of the loss values,
• wherein the computation instruction includes one or more operation fields and an operation code, and the computation instruction includes at least one of:
• a CONFIG instruction configured to configure each constant required by computation for a present layer before computation for each layer of the artificial neural network is started;
• a COMPUTE instruction configured to complete arithmetic logical computation for each layer of the artificial neural network;
• an IO instruction configured to implement reading-in of input data required by computation from an external address space and storage of the data back into an external space after computation is completed;
• a No Operation (NOP) instruction responsible for clearing microinstructions in all microinstruction cache queues presently loaded inside and ensuring all instructions before the NOP instruction are completed, where the NOP instruction does not include any operation;
• a JUMP instruction responsible for enabling the controller to jump an address of a next instruction to be read in the instruction cache unit to implement jump of a control flow;
• and
• a MOVE instruction responsible for moving data of a certain address in an internal address space of the device to another address in the internal address space of the device, where this process is independent from a computation unit, and no resource of the computation unit is occupied in an execution process.